Sentiment analysis of the attorney general's office performance in handling corruption cases on twitter using naïve bayes classification algorithm

Abstract

Corruption is defined as illegal activities such as bribery, fraud and forgery carried out through the abuse of power by public or private officials for personal, financial or other merits. In Indonesia, the Attorney General's Office is one of the institutions that has the authority to handle corruption cases. As the result of the overall business process, public perception is very important. One method to assess public perception is using data collected from social media. Among the many social media, Twitter is known for its high public interaction which then can be used to describe direct people's perceptions. This research aims to create a machine learning model using the Naïve Bayes Classification Algorithm based on Twitter data to determine public sentiment on the Attorney General's Office performance in handling corruption cases. As for the results, we managed to create a model with accuracy, recall, precision, and f-measure values of 74.34%, 71.80%, 73.09%, and 72.44% respectively. From the sentiment analysis result, it can be concluded that the public gives more positive sentiment to the Attorney General's Office in handling corruption cases carried out in the period January 2022 to December 2022 with a percentage of positive sentiment is 61.32% and 38.68% for the negative sentiment.